For 10½ hours, this address repeated the same two-call perpetual settlement attempt every 38 seconds. Across its complete frozen-history record, it made 934 transactions; all completed without an execution error, and every one used the same ordered sequence: authenticate a price-feed update, then pass matched signed orders to a perpetual exchange’s settlement path.
That repetition is consistent with machine-like periodic execution. It does not establish who controlled the address, whether either order filled, or whether the activity was profitable.
A remarkably steady rhythm
The complete-history record runs from 13:02:16 to 23:37:23 UTC on 2026-07-19. The median gap was 38.645 seconds, the mean was 40.844 seconds, and the 90th-percentile gap was 74.953 seconds. The longest observed idle period was 128.666 seconds. The busiest minute contained six transactions.
The repertoire never visibly changed during the window: every hourly bucket was led entirely by execute-paired-order-settlement. The detailed transaction stream is a separate, capped view containing the latest 500 transactions, from 17:52:57 to 23:37:23 UTC; all 500 had the same behavior and pattern as the complete history.
One act, repeated 934 times
The reusable explanation says the first call authenticates a price-feed update and supplies market context; the second passes matched signed orders into paired settlement. The transaction shape confirms the exact calls as price_feed::update_price_feed followed by exchange::trade_v3.
Three checkpoints show how little the script varied:
AsBXVFCu, the first transaction at 13:02:16, recorded a sender-bound change of −0.007202664 SUI.5yREjUe5, the largest absolute gas transaction, ran at 18:03:16 and recorded −0.00740052 SUI.FVDtcCvx, the final transaction at 23:37:23, recorded −0.007201512 SUI.
Each carried the same two-step narrative. In each case, reaching the order-fill call is evidence of an attempted settlement path, not proof that an order filled.
The money story is mostly execution cost
Across all 934 transactions, the recorded sender movements are SUI outflows associated with gas and the sender’s pre-transaction holding balance. The complete-history gas total was 6.71730164 SUI. The matching holding-balance figure is the same accounting movement represented from another perspective and must not be added to the gas total. Movement coverage reported no omitted sources.
No application-asset inflow or outflow is established in the sender-bound movement record. That leaves a bounded conclusion: this address repeatedly paid for settlement-path attempts, while the available evidence does not quantify a filled position, received proceeds, fees, or profit.
The same address paid its own gas in all 934 transactions; no sponsored transactions were recorded. That identifies the paying on-chain address, not a person, company, operator, or strategy owner.
What remains unresolved
The stable behavior resembles scheduled automated settlement activity, supported by its uniform two-call shape, near-constant cadence, and unchanged repertoire across the frozen window. It could still reflect a different process with the same transaction structure. The evidence establishes repeated price-feed refreshes and settlement-path executions; it does not establish successful fills, economic benefit, or the identity behind the address.